Jihwan Jeong

Title: Jihwan Jeong: Innovator in Online Class-Incremental Continual Learning

Introduction

Jihwan Jeong is a notable inventor based in Toronto, Canada. He has made significant contributions to the field of machine learning, particularly in the area of online class-incremental continual learning. His innovative approach addresses the challenges of preserving knowledge while adapting to new data.

Latest Patents

Jihwan Jeong holds a patent titled "Online class-incremental continual learning with adversarial Shapley value." This method focuses on scoring training data samples to maintain latent decision boundaries for previously observed classes. It promotes learning from new images in an online data stream. The process involves receiving new input batches, retrieving relevant data from memory, and updating memory to retain previously learned knowledge. Additionally, the method employs data valuation based on KNN-SV for strategic data selection.

Career Highlights

Throughout his career, Jihwan Jeong has worked with prominent organizations, including LG Electronics Inc. and the University of Toronto. His experience in these institutions has allowed him to refine his skills and contribute to groundbreaking research in machine learning.

Collaborations

Jihwan has collaborated with talented individuals such as Dongsub Shim and Zheda Mai. These partnerships have fostered innovation and enhanced the development of his research projects.

Conclusion

Jihwan Jeong is a pioneering inventor whose work in online class-incremental continual learning is shaping the future of machine learning. His contributions are invaluable to the field, and his innovative methods continue to inspire advancements in technology.